XVI. Education

Ke-young Chu, and Richard Hemming
Published Date:
September 1991
  • ShareShare
Show Summary Details

Why does the government have a dominant role in the provision of education?

How should education expenditure be financed? What is the scope for user fees?

Can the efficiency of public education provision be improved?

Education expenditure at the general government level typically represents from 4.5 percent to 7 percent of GDP in industrial countries. In developing countries, education expenditure ranges from 2.5 percent to 7.5 percent of GDP, and represents a smaller share of GDP on average than in industrial countries. As a share of total expenditure at the general government level, education expenditure often exceeds 10 percent.

Most governments are direct providers of educational services on a large scale, and at the primary and secondary levels, at least, their share of the total supply of such services typically exceeds 90 percent. In addition to this direct role in the market for education, governments influence the demand for educational services by their provision of scholarships, grants and living allowances for students, through decisions regarding fees and tuition charges, and by the regulatory framework they establish for private education. Because public schools and colleges are often administered at the regional or local level, and because the local revenue base may not be sufficient to finance expenditure requirements, the government’s role in education raises important issues concerning intergovernmental financial relations and coordination (see the note on Transfers to Local Government for a general discussion of these issues).

Objectives of Education Expenditure

The dominant role governments tend to play in the market for education reflects not only the importance governments have attached to what is taught but also special economic aspects of education.

Education expenditure as an investment

Expenditure on education may have a large element of consumption, but its character as an investment in human capital is especially important. One special feature of an educational investment is its long gestation period: thus, the expansion of the primary school system does not yield its returns until after 6-8 years; a graduate of medicine represents an investment of perhaps 20 or more years of formal education. The long gestation period required for education to pay off and the limited opportunities to borrow against prospective human capital imply that education costs are usually entirely or substantially subsidized. This also explains why certain kinds of vocational and professional training are provided by employers through apprenticeship programs.


Education also offers wide-ranging external benefits to society. By improving the quality of the work force, education increases the rate of return on other investments and promotes growth; education provides the scope for greater specialization of labor and can facilitate more outward-looking economic development; and education is viewed by many governments as a means of promoting national identity, loyalty, and generally more social patterns of behavior. Because these external benefits do not accrue entirely to the student, there is a further rationale for the government to subsidize at least part of education costs. However, governments tend to supply educational services directly, rather than subsidize the operations of private schools. It is also worth noting that many of the objectives of public policy in the educational field could be achieved indirectly through subsidies and grants to the private sector rather than through the direct provision of educational services. But the direct provision of educational services allows governments to exercise more control over curriculum content and educational standards, and may therefore be viewed as a relatively efficient form of regulation.

Evidence on the rate of return to education

One of the most important conclusions to be drawn from the substantial body of literature on the economic aspects of education is that the returns to investment in education are generally high, particularly in less developed countries where the supply of educated persons is relatively small. These studies typically take into account both the direct cost of education—that is, teachers’ salaries, the cost of classroom aids and other educational materials, and capital costs—as well as the indirect costs—mainly the income forgone by the student during the period spent in school or college; they do not as a rule attempt to measure the benefits of education other than the higher income that is typically associated with an increase in the number of years of formal education. The private rates of return to education are always higher than the social rates of return, because the private calculation excludes that part of the cost of education financed by the government. Another common finding is that the social rate of return to primary education is greater than that to higher education, because costs per student of higher education are typically a large multiple of the costs per student of primary education. In anglophone Africa, for example, the unit cost of higher education has been estimated to be 50 times that of primary education—see Table 1.

Table 1.Public Expenditure per Student on Education and Enrollment Ratios, Major World Regions, around 1980
Expenditure per

Student as a Multiple

of Primary Level

Enrollment RatioNumber of

(In percent)
Developing countries2.926.47523774
Anglophone Africa2.851.17717116
Francophone Africa4.927.74614218
South Asia2.314.9711944
East Asia and Pacific1.810.7874396
Latin America2.99.890441219
Middle East and North Africa14.075.08236911
Industrial countries1.12.2100802120
Source: World Bank.
Source: World Bank.

The incidence of education expenditure

The incidence of education expenditure—that is, the relationship between expenditure per family and family income—can vary quite substantially from one level of education to another. Thus, at the primary level, public expenditure may have a progressive incidence, because expenditure per pupil is constant regardless of the level of income of the pupil’s family. This observation needs to be qualified, however, to the extent that the enrollment ratio rises with the level of family income, and to the extent that local governments spend more on students from affluent neighborhoods. Students at higher levels of education tend to come from wealthier families, particularly in developing countries, because in spite of the large share of direct costs paid by the government in such countries the private cost of higher education will typically be prohibitive for students from poor families. As a result, a relatively large part of the government’s education budget benefits a small number of persons from wealthier families, which has a significant regressive impact.

Implications for public policy

Capital market imperfections and positive externalities that may be associated with education imply that public intervention in the market for education has the potential both to improve allocative efficiency and to achieve a more equitable distribution of the direct benefits of education. For example, if the direct provision of education by the government, either free of charge or at highly subsidized rates, keeps more students in school for longer periods, and the return to society from further education is sufficiently high, then allocative efficiency is improved. Similarly, the distribution of income can be made more equal if the government’s involvement results in a substantial increase in the enrollment of students from poor families, especially at higher levels.

Financing Education Expenditure

The very substantial costs of universal or widespread public education, and the sizable benefits that education can bestow on the educated in the form of more highly paid and prestigious employment, imply that tuition fees and other charges have a role to play in education financing and in the allocation of resources in the educational sector. Thus, a basic issue that must be addressed is the role of user charges. A related issue is the role of student loan programs in compensating for capital market imperfections.

The role of user charges

Charging fees for public schooling is a highly contentious issue. At the primary level, where such charges might affect a large number of pupils from poor families, the use of fees could have a regressive impact on income distribution and lead to the premature end of schooling. Where practical, these adverse consequences can, however, be avoided through personalized prices (see the note on Pricing and Cost Recovery for further discussion). At the higher level, however, there is more scope for using fees to enhance allocative efficiency, increase the resources for education, and produce a more equitable distribution of the benefits of higher education. Whether tuition charges and other fees are an appropriate policy or not again depends on the feasibility of student loan programs and special assistance for students from relatively poor families.

In many developing countries, both equity and efficiency could be enhanced by increasing the share of costs borne by students and their families, provided that assistance were available for the poorer student. The available evidence suggests that there would be little effect on the demand for higher education. The private rate of return to higher education would remain high, and students from wealthier families would not encounter difficulties in financing the extra expenditure. Despite their possibly adverse impact on income distribution, user fees at the primary level have been used to provide pupils with textbooks and other necessary materials that they would not otherwise receive. In a number of countries, user fees are necessary to correct an imbalance between expenditure on teachers’ salaries and other goods and services. In some countries, because government expenditures on tertiary education are a large share of the education budget, the introduction of user charges, and the reinvestment of the proceeds, could finance a substantial increase in enrollment at the primary level. The increase in the share of cost borne by the student would also tend to weed out the less well-motivated students, as well as inspire the others, because the cost of failure for the student is increased.

Student loan programs

The absence of a credit market for educational services helps explain why higher education in less developed countries is generally the preserve of the wealthier classes, and accounts for the creation of student loan programs in industrial countries and a number of middle-income developing countries. Such programs can take various forms: the government can confine its participation in the market to providing guarantees to private-sector lending institutions, or it can make loans directly. If loans to students have a grant element, the cost of the loan is typically either borne directly by the budget in the form of a subsidy to the lending institution or implicitly in the form of a below-market rate of interest on government loans.

Student loan programs have the potential to increase the participation in tertiary education of students from poor families and, by increasing the competition for places, may improve educational attainment. These programs can have a significant impact on the distribution of the benefits of education even without a grant element because of the imperfections that characterize the human capital market. Such programs either make very great demands on the administrative capacities of governments or require an efficient and well-functioning financial system or both. As a result, the widespread use of such programs has been confined to a small number of countries. Precisely because investments in education do not create physical collateral, loan delinquency rates in these programs are high.

Efficiency in Education

Allocative efficiency

The evidence of higher social rates of return for primary than for secondary education argues for a shift in expenditure away from the secondary and tertiary levels to the primary level. However, this does not imply that universal primary education should be achieved at the expense of other levels of education. The objective should be to equalize social rates return at the margin. Moreover, this is an objective that should be pursued both between and within programs. Thus, it could be the case that there was over-investment in tertiary programs in the humanities, as evidenced by the fact that their chief accomplishment was producing graduates for an already overstaffed civil service, and under-investment in the sciences and engineering; the appropriate policy would therefore be a shift in resources out of humanities at the tertiary level to science and engineering, and to primary education. Interactions between education and other public expenditure programs should also be taken into account. For example, the mix of programs at the higher level also has important consequences for public sector wage and employment policy when there are pressures on the government to be an employer of last resort—see the note on Public Sector Employment.

Productive efficiency

Another conclusion that emerges from the literature is the importance of the input mix in education. In particular, there is evidence in many countries whose budgets have been under strain that expenditure on classroom aids and textbooks at the primary level is well below what it should be. Thus, in some African countries, schools with only one textbook per class are far from uncommon. This situation may result from the fact that, with pressure to reduce funding for education, political pressures cushion teachers’ salaries from the consequences of budgetary austerity. Another possible explanation is the overcentralization of the educational establishment’s managerial process, so that budgetary decisions are not made by those most familiar with students’ needs or by those with the incentive or desire to respond to these needs.

Productive efficiency has been enhanced in some countries by the application of modest user charges (textbook fees) at the primary level, and by the devolution of management to local government. Even at the primary level, user charges have apparently not had a large impact on demand for school places and, if spent on the right inputs, can have a significant impact on the rate of return to primary education. It may be required, however, that user charges be earmarked for expenditure on these particular inputs (see the note on Pricing and Cost Recovery for further discussion).

Country Illustration

Distribution of education benefits in Brazil

The incidence of public education expenditure in Brazil by income class varies remarkably from one level of schooling to another. At the primary level, where enrollment ratios are about 90 percent, children whose family income is in the lowest decile are estimated to have accounted for 15 percent of total enrollment in 1980, but only 1 percent of total income. Children from families in the top two deciles, with 59 percent of total income, accounted for only 5 percent of total enrollment. Unless the value of expenditure per pupil rises quite strongly with the pupil’s family income, Brazil would be spending more on primary education for the poor than for the affluent, and the incidence of expenditure at this level would be highly progressive. At higher levels of education, however, the incidence of public education expenditure becomes less progressive. Thus, at the secondary level, the share of students from poor families falls to 3 percent, and that of students from the most affluent 20 percent increases to 16 percent. At the tertiary level, the degree of progressivity falls even further, because the share of students from the poorest families falls to 1 percent, and the share of the most affluent rises to 48 percent. Overall, however, the incidence of public education expenditure appears to be progressive: expenditure per student rises less rapidly than income. At the same time, there is obvious scope for increasing the degree of progressivity, because the budget for tertiary education absorbs about 45 percent of the total education budget, and its direct benefits accrue disproportionately to the better off.

The Brazilian educational system is also characterized by important regional and institutional disparities in expenditure per pupil at the same educational level. For example, primary-level state schools in the Northeast in 1983 spent, on average, over three times as much per student as municipal schools, and rural schools spent less than urban schools. These disparities reflect the fact that primary education is financed largely at the state and local government level, where the tax base per capita varies substantially across the country, and redistributive grants from the federal government play a relatively limited role.

Although primary school enrollment ratios are high in Brazil, the disparity in expenditures across the country and the high failure rates point to a problem of quality at that level. Moreover, enrollment ratios decline from about 90 percent at the primary level to 20 percent at the secondary level, although evidence suggests that the social rate of return to secondary education is high. Both equity and efficiency considerations suggest that the share of total education resources devoted to the lower education levels be increased. One component of a solution might be the imposition of user charges at higher levels; at the moment, the direct costs of public education are subsidized almost 100 percent.


    Psacharopoulos, George, “Education and Development: A Review,” The World Bank Research Observer, 3, 1988.

    The Financing of Education in Developing Countries: An Exploration of Policy Options (Washington: World Bank, 1986).

    Other Resources Citing This Publication